Punishment as a training technique...

[NotEnoughTime]

Whenever someone posts that they utilize a spray bottle, hand quake, or some other "punishing" technique there are always three or four people that jump all over them saying that they should never do this.

But why?

I've seen quite a few people ask and it is a darn good question....

First we need to define exactly what punishment is.... And then look at why it is that we engage in training that involves what other people say is punishment?

Here is punishment as I understand it: Punishment is a training technique that is used to reduce unwanted behaviour.

Through punishment we are attempting to reduce or stop screaming, feather plucking, biting and so on, in our fids. Through the threat of punishment we attempt to have the bird stop whatever it is they are doing. For instance if a spray of water is used to reduce screaming then lifting up the spray bottle with often result in screaming stopping - if, indeed, the spray of water works as a "punishment."

There are two types of punishment if you accept the definition I presented. Positive punishment which is the application of a stimulus (a spray of water) to reduce a behaviour (screaming) and Negative punishment (a time-out) to reduce a behaviour (screaming). Negative punishment involves removal of a positive stimulus while positive punishment involves the appication of a negative stimulus.

The difference between positive and negative punishment is quite important.

There are several issues that occur with positive punishment. Application of a stimulus can lead to fear, avoidance and aggressive behaviours. For example if we are using a water spray to decrease screaming then a possible response could be avoidance of the water bottle used for spraying, avoidance of the individual doing the spraying, or aggression towards the bottle or the person holding the bottle.

Negative punishment, on the other hand, does not have a direct stimulus and therefore is less likely to end up with undesirable behaviours. One has to be careful with negative punishment technique. A time-out is a well-known negative punishment technique, but if one "removes" the bird from a stimulating environment to a neutral environment using the hand (or a finger) then the hand (or finger) can become the punishing stimulus and biting or avoidance behaviours can result.

Punishment as a training tool is also thought of as a poor training tool because it does not tell the parrot what we want the bird to *do.* As a result it is kind-of a double negative - the bird is hit with a stream of water and it ceases to squack (which, for whatever reason, it likes doing). But what does the parrot *do* instead? How does the bird know what to do? All it knows is it was just squirted out of the blue... And if it doesn't like it it will attempt to avoid such squirts in the future by doing whatever it can to avoid the squirt.

If the parrot learns that the squacks are resulting in the squirt then as soon as the water bottle is lifted the squacks will cease. However, if the parrot does not associate the squacks with the squirt then other behaviours could occur... Attempting to dodge the squirt by moving to another location (while still squacking perhaps) for example...

Once learned, avoidance behaviours are extremely resistant to being "untrained." So once your fid has developed a "dodging" behaviour then it is a long, slow process to "untrain" the behaviour. Just ask people who "chase their birds" around the house to get them back to the cage.

So why do we use punishment?

We do it because it is effective and rewarding *to us* in the short to intermediate term. A spray of water quiets a bird, quaking the hand stops the biting... It is quick and effective and is rewarding to us.

But one has to keep in mind that the purpose of a punishing technique is to *reduce* the behaviour. If the behaviour stops immediately but does not reduce in frequency or actually increases in frequency then the punishment technique is actually a reinforcing technique and the bird is simply being training to scream more often (if for a shorter time) until you give it what it wants (maybe a spray of water).

Punishment techniques have their place. A parrot that is feather plucking or self-mutilating, and no reinforcer can be found that overcomes this behaviour, may be considered for a punishment technique. Or a bird that is super-aggressive and does not respond to positive training techniques may also be considered.

But whenever punishment it is used it *should* be backed-up positive reinforcement training of alternative behaviours so that the bird knows what to do. The aggressive bird may find that head scritches are much better than attacking the nearest person - or the self-mutalating bird may find that flight feathers are kind-of nice to have...

So... The reason why I, and so many others, "jump all over" people who are attempting to apply a punishment technique is because it can lead to severe problems with behaviour in the long run...

Please... think long and hard about it before you attempt a punishment technique on any animal but especially a parrot...

Reference material:

Behavior Analysis and Parrot Learning,
S.G. Friedman, Steve Martin, and Bobbi Brinker

ScienceDirect - Journal of Exotic Pet Medicine : The Art of Training Parrots

Secrets of Parrot Training

[chapala]

Just want to say that I'm a 2004 graduate of Susan Friedman's course and believe in the value of positive reinforcement training. Negative reinforcement intended to decrease certain behaviors is burdened with lots of possible side effects as you mentioned. I am wondering about your statement:

"A parrot that is feather plucking or self-mutilating, and no reinforcer can be found that overcomes this behaviour, may be considered for a punishment technique."

Since a plucking or mutilating bird without medical issues may be feather destructive for psychological reasons, it seems to me that punishment in such a case might increase the stress level for the bird who already may have high stress levels. Since you didn't mention a specific technique you would recommend in such a case, I may not be understanding your meaning here, but it just struck me as a little dangerous to make that statement to people with plucking birds who don't understand training principles.

[NotEnoughTime]

Originally Posted by chapala

Since you didn't mention a specific technique you would recommend in such a case, I may not be understanding your meaning here, but it just struck me as a little dangerous to make that statement to people with plucking birds who don't understand training principles.

Good point...

[NotEnoughTime]

Hmmm - my edit time expired...

I don't recommend punishment as a training technique - neither positive nor negative punishments... I much prefer, and strongly recommend, positive techniques.

What is perceived as a negative technique (a time-out for instance) can actually become a positive punishment technique since the parrot must be removed or the environment changed. The conditioned stimulus of a "finger" can be paired with the negative punishment (a time-out) and becomes a conditioned punisher. This can lead to avoidance behaviours that can be very hard to change.

However, in order to get a parrot to simply *stop* a behaviour, a positive or negative punishment technique may have to be used. This is a situation of last resort in my opinion... Not something I would normally or even occasionally do... Rarely it has happened that I've had to resort to this technique... But when I have it has always been combined with a positive technique - so chewing on a toy is rewarded while the chewing on feathers stops. Perseverative behaviours are extremely hard to correct and are very resistant to extinction - and sometimes there is simply no other way to stop the behaviour and this is the *only* time that a negative reinforcement technique makes sense IMO.

[chapala]

Yes, but I thought your statement above could be dangerous since it might lead people who aren't versed in positive training techniques to squirt, yell or worse at the plucking, biting or screaming bird.

Differential reinforcement of an incompatible behavior (DRI) is a tool commonly used by positive reinforcement trainers to work on extinguishing undesirable behavior.

Certain behaviors are very difficult to change, and punishment can seem like a logical choice to a frustrated owner. The side effects of punishment however can damage severely the relationship between bird and owner.

[Abby]

I'm doing Susan's course atm and after only the second lecture it's really making me think as to how I'm managing my different birds.

Before I started the course I'd started to look at Arthur refusing to go back in his cage. Yep the classic chase the birdy thing. It was getting us nowhere and was causing problems in our relationship. Nowadays when I decide it's time for Arthur to go back I show him and almond, put them in his food dish in the cage making sure he's watching and sure enough, he goes back in there no fuss. He's getting something out of going back in his cage (the almond) which seems to be more desirable then spending extra time out.

[NotEnoughTime]

Originally Posted by chapala

Differential reinforcement of an incompatible behavior (DRI) is a tool commonly used by positive reinforcement trainers to work on extinguishing undesirable behavior.

This process utilizes what is called the matching law in behavioural sciences. Matching says that a behaviour that obtains more "reinforcement" is more likely to be chosen over a behaviour that has less reinforcement.

If a behaviour is chosen to reinforce that is incompatible with a behaviour that you want to reduce or eliminate then the behaviour should be replaced with the trained behaviour. It is important to remember that this process does not extinguish a behaviour - it replaces the behaviour.

Again - there are times (very few and very far between) where no appropriate reinforcer can be found for an incompatible behaviour. That being said I by no means encourage people to use punishment as a training technique...

[chapala]

I do the same with my Grey - he will do just about anything for an almond or other nut! Sure is nice to figure out ways to avoid behavior problems, both the bird and person win.

[searcherangel]

The too that I have when he gets kind of wild will put him back in his cage for five minutes and he calms right down wants to spend time doing the cuddle thing not in the cage . Hope its not a bad thing to do .

[kdd2411]

How exactly do you apply postive reinforcement? For example, my 2 year old African Grey Timneh likes to crawl down from his cage and run over to the couch. He is not allowed to, but will test us all day long. Also, sometimes he makes too much noise, I know a bird is supposed to make noise and we sure let him. However, we can't seem to watch movies when he gets in these moods. The last major problem, is that he won't stop doing a high pitched whistle/scream when we leave the room to go upstairs. We would leave him on the cage, but he has proven we can't trust him by himself. How would I apply positive reinforcement in these situations? Currently, the only thing he listens to is the water bottle and I had no idea it is so bad for them, so I would like to find other (positive) ways to train him and prevent future problems.